Hara Ra wrote:
> OTOH, the process DID work. The sniffing resulted in demoloshing the
> credibility of the putative cowgirl. The next step up is to incorporate
> cryptographic methods, such as signature verification, zero info proofs and
> so on. This could be automated and operate in a manner similiar to the
> discussion and annotation system available at Foresight....
yup. another approach: try to find the types of questions that folks wouldnt mind revealing about themselves, and no one would have much motive to wreck your database. the example of the link between coffee and kidney stones has stumped doctors, because you have opposing effects. caffiene is a diuretic, so it stimulates urination, so it encourages thirst and increased drinking of fluids: good, however coffee also contains chemicals that may produce oxalates: bad. so is coffee good or bad for kidney stones?
nowthen, most folks dont mind telling how many kidney stones they have suffered, nor do they mind telling how many javas they quaff, however, for these data to be interpretable, we would also need to know... the age of the respondant. {8-[ which people dont like to tell.
so. im still looking for a good test case. spike